Accuracy of Bilistick (a Point-of-Care Device) to Detect Neonatal Hyperbilirubinemia

Abstract

Abstract Introduction Early diagnosis and appropriate management of neonatal jaundice is crucial in avoiding severe hyperbilirubinemia and brain injury. A low-cost, minimally invasive, point-of-care (PoC) tool for total bilirubin (TB) estimation which can be useful across all ranges of bilirubin values and all settings is the need of the hour. Objective To assess the accuracy of Bilistick system, a PoC device, for measurement of TB in comparison with estimation by spectrophotometry. Design/methods In this cross-sectional clinical study, in infants who required TB estimation, blood samples in 25-µl sample transfer pipettes were collected at the same time from venous blood obtained for laboratory bilirubin estimation. The accuracy of Bilistick in estimating TB within ±2 mg/dl of bilirubin estimation by spectrophotometry was the primary outcome. Results Among the enrolled infants, 198 infants were eligible for study analysis with the mean gestation of 36 ± 2.3 weeks and the mean birth weight of 2368 ± 623 g. The median age at enrollment was 68.5 h (interquartile range: 48–92). Bilistick was accurate only in 54.5% infants in measuring TB within ±2 mg/dl difference of TB measured by spectrophotometry. There was a moderate degree of correlation between the two methods (r = 0.457; 95% CI: 0.339–0.561, p value < 0.001). Bland–Altman analysis showed a mean difference of 0.5 mg/dl (SD ± 4.4) with limits of agreement between −8.2 and +9.1 mg/dl. Conclusion Bilistick as a PoC device is not accurate to estimate TB within the clinically acceptable difference (±2 mg/dl) of TB estimation by spectrophotometry and needs further improvement to make it more accurate.